MOUSING WITH SPSS Frances Provan, Information Services, Edinburgh University Useful point and click

MOUSING WITH SPSS

Frances Provan, Information Services, Edinburgh Frances Provan, Information Services, Edinburgh UniversityUniversity

Useful point and clickUseful point and click

ASSESS YORK 2007 Mousing with SPSS 2

All at sea


There's a lot in SPSSThere's a lot in SPSS


Today’s blubber coversToday’s blubber covers

Things I like…Things I like… On the fly utilitiesOn the fly utilities WizardsWizards Open secretsOpen secrets Old SPSS favouritesOld SPSS favourites

All by mouseAll by mouse


SPSS Versions

Version comparison list from SPSS

http://www.spss.com/software_version/ Lists changes between versions

and new features. Goes back to version 6

What’s new for latest version

http://www.spss.com/spss/whats_new.htm

http://www.spss.com/software_version/






Some slide shorthand

{version no. first appeared} {14} introduced in version 14

Menu pathsMenu > Sub Menu > Sub menu e.g. File > Open > Data…

i.e. choose the File menu, select Open from the File menu, then select Data... from the Open submenu

Plus point Plus point something I like…


Did you know you could…

Define your own styles? Create charts from tables? Web pages from results? Document your data files? Get previews when building

chart and tables?


Nice newish graphs…

Dot plots (about time too!) {13}Graphs > Scatter/dot…

Then choose Simple Dot Population pyramids {13}

Graphs > Population Pyramid Panels for ordinary graphs {13}


Dot plot


Population Pyramids


Population Pyramids: categories

Each category -a separate pane 3 level variable


Panels

Lovely idea… 1st in interactive plots Now Rows and columns {14} Or Multiple factors

Separate or nested

Great for comparisonsGreat for comparisons


Dot plots: row panels


Dot plots: Column Panels

Panelled in columns


Dot plots: Row and Column Panels


Population Pyramids: Row Panels

Single Row factor


Your own style Pivot tables & Table looks Interactive Charts & Chart looks Charts & chart templates Edit graph/chart/table to access

i.e. Double click on it Change your style with

Edit > Options… Charts, Interactive or Pivot Tables tab


Pivot tables

Double click to edit Change the look

Format > Tablelooks… Change rows, columns & layers

Pivot > Pivoting Trays Keep dimension changes

Pivot > Bookmarks


Tablelooks Default 195 98 2

45% 20% 5%

218 338 12

50% 68% 30%

21 61 26

5% 12% 65%

Very Happy

Pretty Happy

Not Too Happy

GeneralHappiness

Exciting Routine Dull

Is Life Exciting or Dull



195 98 2 Very Happy

45% 20% 5%

218 338 12 Pretty Happy

50% 68% 30%

21 61 26

General Happiness

Not Too Happy

5% 12% 65%

Academic


Chart from a table

Double click on the table Right click, Create Graph > Bar Be careful what is selected

Total lines look daftUse layers to be selective


Bar chart from table

General Happiness Very Happy

General Happiness Pretty Happy

General Happiness Not Too Happy

Row

Is Life Exciting or Dull ExcitingIs Life Exciting or Dull Routine

Is Life Exciting or Dull Dull

Column

0

100

200

300V

alu

es

Statistics : Count


Interactive graphs {long time}

Graphs > Interactive > Bar A lot of the ‘new’ graph features

already there Panel variables Chart looks Exploratory data analysisExploratory data analysis Mmm.. Leopard skin Mmm.. Leopard skin

barcharts..barcharts.. Tastefully tacky


Grrraphs - to the leopard skin

Double click to open Select object

right clicking on bar select all bars

Choose Fill button Pattern .bmp image

Or choose Format >

chart properties Filled Objects tab


Chartlooks & Chart templates

Available when editing Interactive Graphs

Format > Chartlooks… Chart builder

File > Save Chart Template File > Apply Chart Template

Use your own style by default Edit > Options


Grrraph to Dante


Builders

Custom tables {12?} Analyze > Tables > Custom Tables..

Chart builder {14} Graphs > Chart Builder…Still newNot sure I like them yet…


A Customized table

With academic table look…



195 98 2 Very Happy

45% 20% 5%

218 338 12 Pretty Happy

50% 68% 30%

21 61 26

General Happiness

Not Too Happy

5% 12% 65%


A built stacked bar chart


Exporting Output

File > Export Export as:

HTML - web pages {7} Anyone can read it ‘clean’ HTML separate picture files

Word {11.5} All in one file Big files.

Powerpoint (see later) {13} Excel (tables only) {11.5} PDF {15}

No copy and pasteNo copy and paste


Controlled output export Can specify:

Amount of output: All output (includes hidden stuff) All Visible output Only selected objects

Charts, documents or both Image types for charts Output types, as above

Use in conjunction with OMS {14} Utilities > OMS Control Panel Utilities > OMS Identifiers


Export to MS Powerpoint {13}

Not everything translates, but you get:Pivot tablescharts {14}MapsTrees

Used it for some slides…


Case Processing Summary

Cases

Valid Missing Total N Percent N Percent N Percent General Happiness * Is Life Exciting or Dull * Region of the United States

971 64.0% 546 36.0% 1517 100.0%

Table straight to powerpoint


Crosstabs output for:

Analyze > Descriptive Statistics > Crosstabs

Variables Row is happy Column is life Layer variable is region

Count & column statistics from Cells Chi-squared tests from Statistics Ticked Clustered bar chart


Case Processing Summary

Cases

Valid Missing Total N Percent N Percent N Percent General Happiness * Is Life Exciting or Dull * Region of the United States

971 64.0% 546 36.0% 1517 100.0%

Case processing


General Happiness * Is Life Exciting or Dull * Region of the United States Crosstabulation

Is Life Exciting or Dull Region of the United States Exciting Routine Dull Total

Count 80 40 1 121 Very Happy

% within Is Life Exciting or Dull 43.0% 17.9% 5.3% 28.2%

Count 99 159 6 264 Pretty Happy


Count 7 25 12 44

General Happiness

Not Too Happy


Count 186 224 19 429

North East

Total






Count 4 23 8 35

General Happiness

Not Too Happy


Count 107 145 12 264

South East

Total



% within Is Life Exciting or Dull 44.7% 19.5% .0% 31.7%



Count 10 13 6 29

General Happiness

Not Too Happy


Count 141 128 9 278

West

Total


Crosstabs table


Chi-Square Tests

Region of the United States Value df

Asymp. Sig. (2-sided)

Pearson Chi-Square 94.279(a)

4 .000

Likelihood Ratio 70.060 4 .000 Linear-by-Linear Association

59.027 1 .000

North East

N of Valid Cases 429

Pearson Chi-Square 52.875(b)

4 .000


37.470 1 .000

South East

N of Valid Cases 264 Pearson Chi-Square 51.772(

c) 4 .000


29.150 1 .000

West

N of Valid Cases 278 a 1 cells (11.1%) have expected count less than 5. The minimum expected count is 1.95. b 2 cells (22.2%) have expected count less than 5. The minimum expected count is 1.59. c 2 cells (22.2%) have expected count less than 5. The minimum expected count is .94.

Chi-squared tests


Barchart 1


Barchart 2


Barchart 3


On the fly

Visual bander {12} Transform > Visual Bander

Define Variable properties Data > Define Variable Properties

Copy data properties Data > Copy Data Properties

Automatic recode {forever} Transform > Automatic Recode


Visual Bander {12}

Interactive tool to categorise data Cut points

Manually defined Equal ranges Equal counts Using mean and standard deviations

Labelling, either automatic or manual Love combined recode & Love combined recode &

labelling labelling


Define Variable Properties {?}

Easier than Variable ViewEasier than Variable View Use it to:

Type labelsView whole definitionCopy definitions to and from

other variables


Automatic Recode {forever} String to numeric Large numeric codes Doesn’t miss out values Sorts out messy codes Keeps coding to use again Keeps coding to use again

{recent}{recent}


Data file stuff

Showing data file information Display data file information >

external file Display data file information >

Working file Data file comments

Utilities > Data File Comments Multiple open datasets {14}


File Information Source E:\Program Files\SPSS14\1991

U.S. General Social Survey.sav

Type SPSS Data File Creation Date 16-SEP-2002 11:12:10 Label None

Data Type Case N of Lines of Documents 391 Variable Sets Yes Trends Date Information None Multiple Response Definitions Yes

Data Entry for Windows Information None

TextSmart Information None

File Contents

Clementine Information None N of Cases 1517 N of Defined Variable Elements 43

N of Named Variables 43 Weight Variable None

Data Information

Compressed Yes

File information


Variable Information

Name Position Label Measurement Level Format Column Width Alignment Missing Values

sex 1 Respondent's Sex Nominal F1 8 Right

race 2 Race of Respondent Nominal F1 8 Right

region

3 Region of the United States

Nominal F8.2 8 Right

happy 4 General Happiness Ordinal F1 8 Right 0, 8, 9

life

5 Is Life Exciting or Dull Ordinal F1 8 Right 0, 8, 9

sibs

6 Number of Brothers and Sisters

Scale F2 8 Right 98, 99

childs 7 Number of Children Ordinal F1 8 Right 9

age 8 Age of Respondent Scale F2 8 Right 0, 98, 99

educ

9 Highest Year of School Completed

Scale F2 8 Right 97, 98, 99

paeduc

10 Highest Year School Completed, Father


maeduc

11 Highest Year School Completed, Mother


speduc

12 Highest Year School Completed, Spouse


prestg80

13 R's Occupational Prestige Score (1980)

Scale F2 8 Right 0

occcat80 14 Occupational Category Ordinal F8.2 8 Right

tax 15 R's Federal Income Tax Ordinal F1 8 Right 0, 8, 9

usintl

16 Take Active Part in World Affairs

Ordinal F1 8 Right 0, 8, 9

Variable Information


Value Labels

Value Label 1 Male sex

2 Female 1 White 2 Black

race

3 Other 1.00 North East 2.00 South East

region

3.00 West 0(a) NAP 1 Very Happy 2 Pretty Happy 3 Not Too Happy 8(a) DK

happy

9(a) NA 0(a) NAP 1 Exciting 2 Routine 3 Dull 8(a) DK

life

9(a) NA 98(a) DK sibs

99(a) NA 0 0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 Eight or More

childs

9(a) NA 98(a) DK age

99(a) NA 97(a) NAP 98(a) DK

educ

99(a) NA 97(a) NAP 98(a) DK

paeduc

99(a) NA 97(a) NAP 98(a) DK

maeduc

99(a) NA speduc 97(a) NAP

Value Labels


Data File Comments

Just add text to dialog box Time stamped Use comments to:

Describe where data has come from Keep codebook with the data Document changes to data file

Can print to output Documents command is useful Remains with your SPSS Remains with your SPSS

data filedata file


Comments box


The wizards

Date and time {13} Transform > Date/Time

Restructure data {11.5} Data > Restructure…

ODBC & read text File > Open Database File > Read Text Data..

Sample Wizard Analyze > Complex Samples


Date/Time wizard {13}

Date format very useful Do you know how difficult it

used to be to calculate age from date of birth

Loads of things you could only do with syntax before.


Restructure wizard {11.5}

‘long' data files into 'wide' files 'wide' data files into 'long' files cases become variables,

variables become cases Indexing Great for repeated Great for repeated

recordsrecords


Identify Duplicate Cases {12}

Data > Identify Duplicate Cases… Filter out duplicates Create indicator to use elsewhere

E.g. Data > Select Cases… to delete duplicates

Creates indexes One stop shopOne stop shop

easier than sort cases & aggregate


Identify duplicate cases

Indicator of each last matching case as Primary

384 79.0 79.0 79.0

102 21.0 21.0 100.0

486 100.0 100.0

Duplicate Case

Primary Case

Total

ValidFrequency Percent Valid Percent

CumulativePercent


For the enquiring mind

Online help Help > Topics or Tutorial or Case Studies Help button on every dialog box

Help about the procedure Details on choices

Right click (or Mac, Control-click) To see what options you have

Context sensitive menus To get a bit of background


And so much more…

$casenum System variable - Current case order, Copy to record

current order, Sort by copy to return to that order Use with Transform > Compute

Merge/Match Files Merging two data files, IN for case source, BY matches, Data > Merge Files > Add Cases or Add Variables

Count Counts across range of variables, Good for Multiple

Response, Also non response (missing values) Transform > Compute

Frequencies Lists absolutely everything, User & System missing values,

Each separate value Analyze >Descriptive Statistics > Frequencies

Aggregate Checking duplicates (Superceded now), FIRST and LAST

within pre-sorted groups, string variables (command only) Condenses data by any variable, Recently added {14},

Automatic matching in, New datasets in SPSS session Data > Aggregate..

Lag gets next in sequence, Great for selecting duplicates, Use with Data > Sort Cases or Transform > Compute

$casenum System variable - Current case order, Copy

to record current order, Sort by copy to return to that order

Use with Transform > Compute Merge/Match Files

Merging two data files, IN for case source, BY matches, Data > Merge Files > Add Cases or Add Variables

Count Counts across range of variables, Good for Multiple Response, Also

non response (missing values) Transform > Compute

Frequencies Lists absolutely everything, User & System missing values, Each

separate value Analyze >Descriptive Statistics > Frequencies

Aggregate Checking duplicates (Superceded now), FIRST and LAST within pre-

sorted groups, string variables (command only) Condenses data by any variable, Recently added {14}, Automatic

matching in, New datasets in SPSS session Data > Aggregate..

Lag gets next in sequence, Great for selecting duplicates, Use with Data > Sort Cases or Transform > Compute

$casenum System variable - Current case order, Copy to record current order, Sort

by copy to return to that order Use with Transform > Compute

Merge/Match Files Merging two data files, IN for case source,

BY matches, Data > Merge Files > Add Cases or Add

Variables Count

Counts across range of variables, Good for Multiple Response, Also non response (missing values)

Transform > Compute Frequencies

Lists absolutely everything, User & System missing values, Each separate value

Analyze >Descriptive Statistics > Frequencies Aggregate

Checking duplicates (Superceded now), FIRST and LAST within pre-sorted groups, string variables (command only)

Condenses data by any variable, Recently added {14}, Automatic matching in, New datasets in SPSS session

Data > Aggregate.. Lag

gets next in sequence, Great for selecting duplicates, Use with Data > Sort Cases or Transform > Compute

$casenum System variable - Current case order, Copy to record current order,

Sort by copy to return to that order Use with Transform > Compute


Count Counts across range of variables, Good for

Multiple Response, Also non response (missing values)

Transform > Compute Frequencies

Lists absolutely everything, User & System missing values, Each separate value

Analyze >Descriptive Statistics > Frequencies Aggregate





$casenum System variable - Current case order, Copy to record current

order, Sort by copy to return to that order Use with Transform > Compute


Count Counts across range of variables, Good for Multiple Response, Also non

response (missing values) Transform > Compute

Frequencies Lists absolutely everything, User & System

missing values, Each separate value Analyze >Descriptive Statistics >

Frequencies Aggregate





$casenum

System variable - Current case order, Copy to record current order, Sort by copy to return to that order Use with Transform > Compute

Merge/Match Files

Merging two data files, IN for case source, BY matches, Data > Merge Files > Add Cases or Add Variables

Count

Counts across range of variables, Good for Multiple Response, Also non response (missing values) Transform > Compute

Frequencies Lists absolutely everything, User & System missing values,

Each separate value Analyze >Descriptive Statistics > Frequencies

Aggregate Checking duplicates (Superceded

now), FIRST and LAST within pre-sorted groups, string variables (command only)




$casenum System variable - Current case order, Copy to record current order, Sort by copy to return to

that order Use with Transform > Compute


Count Counts across range of variables, Good for Multiple Response, Also non response (missing

values) Transform > Compute

Frequencies Lists absolutely everything, User & System missing values, Each separate value Analyze >Descriptive Statistics > Frequencies

Aggregate Checking duplicates (Superceded now), FIRST and LAST within pre-sorted groups,

string variables (command only) Condenses data by any variable, Recently added {14}, Automatic matching in, New

datasets in SPSS session Data > Aggregate..

Lag gets next in sequence, Great for

selecting duplicates, Use with Data > Sort Cases or

Transform > Compute

Those were my Mousing

Tips


Fallen asleep?Fallen asleep?


Useful urls

ASSESS web sitehttp://www.spssusers.co.uk/

SPSS web site: http://www.spss.com/

help system, on-line manuals SPSS mailing list,

http://www.listserv.uga.edu/archives/ spssx-l.html

